Beamforming and power allocation for MIMO-NOMA cognitive networks

Abstract

As a promising technology to meet the dramatic growing demand for high spectrum efficiency and massive connectivity in the fifth generation (5G) wireless networks, nonorthogonal multiple access (NOMA) has attracted significant research attention in recent years. The key idea of NOMA is multiplexing users on the power domain, and receivers utilize successive interference cancellation (SIC) to cancel a significant part of received interference. Cognitive radio (CR) is another technique to boost the spectrum efficiency by allowing secondary users to dynamically access the spectrum of the primary network. However, employing NOMA in the CR networks complicates the problem because NOMA itself is interference limited. Multi-input multi-output (MIMO) system equips multiple antennas on both the transmitter and the receiver side, by adjusting phases and weights of these antennas, which is so called beamforming, interference can be well managed. It is therefore an important and urgent research task to investigate the beamforming and power allocation design in this scenario. The first part of this thesis considers a scenario where a secondary network located at the edge of a primary network, and both of them adopts NOMA for transmission. We focus on designing a coordinated beamforming method for both networks to manage their interference, then optimize the power allocation scheme to improve system utility. The second part of this thesis considers a scenario where secondary users are pairs of IoT devices equipped with a single antenna, and the primary network serves its users with NOMA. In this part, we focus on the optimization of beamforming vectors to reduce the interference to the secondary devices and encourage them to access the spectrum, and we further optimize a pricing scheme which controls the access of secondary users.